My proposed research is designed to examine how young children acquire the meaning of verbs. Specifically, I will examine four relevant sources of information - referential, syntactic, selectional, and prosodic - as these interact to reveal the meaning of new predicates. Previously available methods have allowed us to reason inferentially from the child's behavior in off-line tests of her comprehension. However, a method has recently emerged that allows us to study aspects of the on-line process of sentence interpretation in young children by tracking their eye movements as they interact with a miniature world. I will use this promising technology to see how the young learner goes about extracting the meanings of previously unknown verbs from their complex embedding in the referential and linguistic stimulus situation. My question is how the interpretive process works when the learner is young, and the verb item is unknown. The emerging parsing process, apparently incomplete even in 5 - 7 year old children, can be expected to interact critically with other aspects of the lexical learning machinery.